AWS Certified AI Practitioner (AIF-C01) · Objective 4.2
Approving AI changes
4.2Standard
Change model, data, prompt, and intended use.
Approve based on impact.
Choose two practices.
Hint 1: Focus on the requirement
Record model or data changes with impact, evaluation, approval, and rollback information before release.
Hint 2: Key terms
Review AI change management.
Change management
A controlled process for proposing, reviewing, releasing, and tracking changes.
Impact assessment
An analysis of affected users, data, controls, quality, and operational risk.
Approval gate
A required review decision before a change can proceed to the next stage.
Rollback plan
A tested procedure for returning to a known safe version after a failure.
× Incorrect
Record change, scope, evaluation, risk, rollback, and approver, and require reassessment, stakeholder review, and staged rollout for high-impact changes.
Detailed explanation
A
CorrectRequire reassessment, stakeholder review, and staged rollout for high-impact changes
Important changes receive stronger control.
B
IncorrectApply production changes with no approval record
Auditability is lost.
C
CorrectRecord change, impact scope, evaluation, risk, rollback, and approver
Decisions remain traceable.
D
IncorrectPrompt changes need no approval because they are not code
Behavior and safety can change.
Check official documentation
Compare the following points with the official documentation.
Path & Pass is an independent learning site and is not affiliated with any certification body. Questions are original practice content based on published objectives.